Combined Measures toward Restoration of Temporal Information from Offline Handwritten Text

Abstract

The research described in this paper intends to restore temporal information out of offline handwritten words. Such a system can be a backbone to a number of other more complicated systems such as signature verification, and handwriting analysis. We combined energetic measure with vision rules to estimate the original handwriting. Experimental results indicate a 28% improvement in the estimation accuracy of the combined measure over that of the energetic measure alone.